With current language technologies, localisation of content across various financial channels need not involve a manual workforce that incurs resources, time, cost, and effort.
During the pandemic, businesses and government organizations in India saw a huge language gap while reaching out to people on digital platforms. Sectors like banking, healthcare, education, and so on, saw growing needs for engagement, especially during the lockdown. The rural internet usage in India surpassed that of the urban long before the pandemic hit. Covid-19 has accelerated internet adoption among Indian-language users. When banks, insurance, and fintech companies communicate off the internet with this new, previously overlooked Indian-language customer base, they use local languages. However, the buck stops at digital platforms offering services in a language this new customer base couldn’t comprehend.
A KPMG report states that 72% of Indian internet users cited limited comfort in accessing online content as a barrier to using digital payments. This is a real problem. Imagine living like a tourist or a foreigner who cannot understand the region’s language in your own country!
Localising end-to-end communications with language technologies
One reason banks and fintech institutions have likely shied away from localizing offerings is the sheer scope of the task. There are numerous points of contact with users across various channels, and it can appear a daunting prospect to translate this huge volume of content manually, in multiple languages at that. With current language technologies, localisation of content across various financial channels need not involve a manual workforce that incurs resources, time, cost, and effort.
AI-based localization services can automate most of these processes, doing most of the work over large volumes of text. This is smart localization as opposed to overly complicated manual translation. The need to have large teams and background infrastructure just for localized sites is removed, and human interaction can be limited to quality checks done by specialized translators with domain expertise instead, while all the backend work of creating new localized pages can be handled through automation.
AI helps with scale, speed, and accuracy. Large volumes of text can be translated rapidly using AI, which are then uploaded to the platform and made accessible by users. This localization process should be end-to-end across channels with no break in user experience, meaning that someone using a site in Gujarati should not receive an OTP SMS in Hindi or a payment gateway page in English.
Localized chatbots can prompt users by asking them what it is that they need assistance with through popup chat windows, which users can respond to in their preferred language using built-in Indic input. Depending on the user’s preference, such interactions could be either text or voice.
Overcoming literacy barrier with Indian-language voice technologies
According to a report by WatConsult, 76% of Indian smartphone users are familiar with voice technology. Text-heavy interfaces could also prove challenging for unlettered Indian-language users and those unfamiliar with common UI/UX practices. Voice technology offers a convenient way for banking platforms to offer services to users by speaking to them, prompting responses, and eventually actions and transactions.
Voice can also drive users to interact more closely with platforms. According to a 2020 Google study, 60% of Indian smartphone users interact with and use voice-powered assistants for a range of needs. These actions are 40 times likelier to be action-oriented, clearly showing that voice makes it easier to avail of services.
Indian-language voice bots on the rise
Localized voice bots, in particular, are extremely effective in communicating with Indian language users not used to text-heavy interfaces. Easy to operate, conversational, and accessible, voice bots can literally speak to users and do tasks for them, all in their own language.
First, an automated introduction prompts the user for their requirements and details in their preferred Indian language, offering them the range of services available. Each response by the user to prompt results in further prompts or actions from the bot’s side. For example, a user can check their FD balance just by asking the voice bot “how much money is in my FD today” in Tamil, and the bot will respond with the number in Tamil.
Local-language IVRs automate customer communications
A call placed to a customer care number will prompt users with a prerecorded helpline message asking them to choose their preferred language. After this, it will ask the user to input or speak certain details out loud. Once the user finds their requirement, a confirmation will result in the information being relayed back, all in the user’s originally selected preferred language.
An IVR solution has less interactivity than a voice bot but it can still be especially useful for providing users with details about financial services, or checking account details like one’s bank balance or loan status. In fact, such essential services can even be placed higher up in the chain of queries, ensuring that it takes fewer steps for a user to get the information they need.
Essentially, processing a simple IVR transcript even in English takes about 72-96 hours through the conventional process of studio recording by a voice-over artist. The turnaround time of processing and translating the voiceover into different languages could be reduced to under 3 hours using language technologies, without compromising the naturalness of the human voice regardless of gender.
Broadly, there are three important pillars for digital inclusivity in India.
- Access to affordable smartphones
- Enablement of digital infrastructure
- Availability and access to information in Indian languages
Currently, India is getting comfortable with the first two pillars with the proliferation of smartphones and affordable data infrastructure in most locations. When it comes to the third pillar, robust language technologies and products are the way forward. Being one of the essential services providers, the BFSI industry should step up to cater to the rising Indian-language user base setting the way for other industries and businesses to follow suit.
Arvind Pani is CEO and Co-Founder of Reverie Language Technologies. Views expressed are the author’s personal.